THE RATIO OF THE SPECIFIC HEATS OF GASES. 283 



peculiar relation of the two chlorine atoms to each other 

 when they come in the same molecule is not a result of 

 their being no longer monovalent, but is — according to the 

 most reasonable theory of valency, the electrical theory — the 

 result of their carrying more than one unit of electricity each. 



These are mere speculations at present, but they serve 

 to show the kind of information that may be got by further 

 determinations of the ratios of the specific heats of suitably 

 chosen compounds. It would be interesting, for instance, 

 to investigate a number of compounds containing well- 

 marked radicles, such as NH 2 or OH, to find whether the 

 radicle always contributes the same amount to /3, and also 

 to find whether it is a general law that isomers have the 

 same y. 



Except for the difficulty in making the determination y 

 might be expected to be as serviceable in fixing constitu- 

 tions as the numerous constants, index of refraction, surface 

 tension, etc., already largely used for the purpose by organic 

 chemists. 



An interesting piece of work has been done on nitrogen 

 peroxide by E. and L. Natanson. These physicists deter- 

 mined the ratio of the specific heats over a long range of 

 pressures, simultaneously making careful determinations of 

 the density. They found that at the highest pressure used, 

 when according to the usual theory 1 5 per cent, of the gas 

 was dissociated into N0 2 , y had the value 1*17, which is 

 near the average value for a six atom molecule ; whilst, 

 when the gas was fully dissociated, the value of y was 1*31, 

 about the same as has been found for most three atom 

 molecules, thus showing conclusively that the gas is really 

 dissociating, and that the variation in density is not merely 

 an abnormal case of the usual deviation from Boyle's law. 



The effect of change of temperature on y has not 

 received the attention it deserves. Several observers, 

 notably Wlillner and Strecker, have looked for a de- 

 pendence of 7 on the temperature, but no one has ever 

 found it except Wiillner, who found that the y's of carbon 

 dioxide, carbon monoxide, nitrous oxide, ethylene and 



ammonia all fell as the temperature rose, in some cases by 



20 



